Lab
Écoute ça is a music production laboratory. Each track starts from an idea, research into a musical world, trials, selections and countless tweaks. AI is one of the tools used, but it erases neither the intentions, nor the choices, nor the production work.
01
Everything starts from a mood, a style or a reference. I listen, I research, and I try to understand the codes of the genre before creating anything at all.
02
I develop the lyrics, rhymes, rhythm and syllables from personal ideas and a creative dialogue with ChatGPT.
03
In Suno, I set the style, lyrics, sonic colour and voice, then generate and compare a great many versions, sometimes up to 200, before settling on a direction.
04
In Suno Studio, I work the track track by track: adding, removing, adjusting or transforming elements, sometimes starting from a hummed idea, adjusting lyrics and rebuilding the arrangement.
05
I use Audacity for effects, sound touch-ups, level control and preparing exports before release.
writing partner
I use it as a creative sounding board to develop an idea, unblock a line or explore several directions. It can suggest text, rhymes or rewordings; I also feed it my lyrics to adjust rhythm, sounds and syllable count. The writing takes shape in that exchange, then I select, rewrite and assemble the final version.
production and exploration workshop
Suno is the heart of my production process. I start from a precise intention — style, mood, voice, lyrics or melody — then explore a great many versions before settling on a direction. In Suno Studio, I then work the track piece by piece: adding, removing, adjusting or transforming elements, sometimes starting from a hummed idea, editing lyrics and rebuilding the arrangement. The tool opens up possibilities; selection, iteration and final choices remain at the centre of my work.
finishing, effects and release preparation
I use Audacity for the final sound touch-ups: adding or adjusting effects, fixing details and preparing files before release. It also helps me control levels and normalise volume so each track stays balanced and isn't degraded by the processing platforms apply. It's a finishing step, between music production and final mastering.
I'm not doing this for the money, or for the glory. I do it for myself first, because a song that exists is worth more than an idea left in a corner. And once it exists, it might as well be useful.
That's why tracks are offered for download as they go: the full MP3, and, when it exists, the instrumental version. If a file is still missing, it simply hasn't been uploaded yet — it's coming.
If you want to sing over it, rap over it, slow it down, break it apart, remake it better: go ahead. Drop the SoundCloud link to your version from the song's page. I'll listen, and if it speaks to me, it gets published alongside the original. Nothing to sign, nothing to pay — just the urge to see where it can go.
Rather than staying vague, every published song states the actual amount of generative assistance it required, with a note explaining what was explored, selected, reworked or redone by hand.
This estimate is not a technical measurement: it's an editorial statement about how much room the tools took up in the final result. It's broken down by role — writer and lyrics, composer and music, producer and arrangements — then averaged into an overall percentage.
No percentage is shown here: they all live on the songs themselves. Open a track's page to read the full credits and the AI share involved, or an album's page to see the percentage for each track.